Three estimates - Different Panels

[bob-n]

"can anyone provide any non-anecdotal, unbiased, evidence-based data on the failure rate / TCO of micro-inverters vs string inverters?"

This report is the most objective thing I've read on the subject.

https://www.mcelectrical.com.au/micro-inverters/

I'll quote a few paragraphs from the report below.

"In the last five years, we have installed about 1550 micros on 77 homes. Most of these systems are over four years old, so they are close to halfway through their ten year warranty period. To date, we’ve had only four microinverter failures (0.25%) and a few communication issues that I’ll address later."

"You could say that 4 out of 77 jobs means about 5% of our Enphase jobs have had a failure. That’s bad compared to a quality string inverter like Fronius. However, that’s not a fair comparison. Enphase’s “decentralised power topology” ensures that if one microinverter fails, you’ll only lose the production of one panel. When a string inverter (like Fronius or SMA) fail, you lose the production of the whole system."

"SolarEdge is a fairer comparison. SolarEdge uses both rooftop power electronics (optimisers) and a central point of failure (the SolarEdge Inverter). How does SolarEdge fair for reliability? 34% of our SolarEdge jobs have had at least one fault. 11% of our SolarEdge inverters have failed so far, taking the whole system down."

"Every solar panel has three bypass diodes to protect them from the shade. If they activate too often because of severe shade, then they may fail, and you will lose 1/3 of the production of your panel. We have only ever picked up six blown bypass diodes from the 5000+ panels that we have installed with individual panel monitoring (Enphase, SolarEdge or Tigo). However we’ve had four Enphase microinverters, and 56 SolarEdge optimisers fail."

My take-away from this report is that Enphase reliability is very good, and much better than SolarEdge optimizers and much better than SolarEdge inverters. Optimizers are on the roof just like microinverters, so still expensive to replace. And any repair job is a hassle, because it requires debug, vendor support, and field work. Don't underestimate the challenge of debug. Every failure is different.

To those who say "hot electronics are unreliable", please understand that it's not that simple. Hot electronics that is engineered for that temperature and uses components that are reliable at that temperature will be reliable. As a simple example, you can buy electrolytic capacitors that are rated for 85C, 105C, 125C, and 150C operation. The difference is materials and construction, not just specs. If the inverter builder uses high-temperature capacitors, they will get much longer life. Yes, the roof can get to 130F (54C) and the inside of the microinverter can get even hotter. But that's still tepid for 125C or 150C components.

Don't forget that the insides of a string inverter is also very hot because it is processing many kilowatts.

Sorry for ranting.

[robbyg]

To those who say "hot electronics are unreliable", please understand that it's not that simple. Hot electronics that is engineered for that temperature and uses components that are reliable at that temperature will be reliable. As a simple example, you can buy electrolytic capacitors that are rated for 85C, 105C, 125C, and 150C operation. The difference is materials and construction, not just specs. If the inverter builder uses high-temperature capacitors, they will get much longer life. Yes, the roof can get to 130F (54C) and the inside of the microinverter can get even hotter. But that's still tepid for 125C or 150C components.

Don't forget that the insides of a string inverter is also very hot because it is processing many kilowatts.

Sorry for ranting.

It's all very relative. A 150C rated electrolytic capacitor will last a lot longer at 50C vs 150C. As you stated the roof can get very hot at 54C, in reality they can get up to 60C and lord knows how much hotter it could be inside that inverter, especially if its sandwiched close to the roof.

Capacitors are not the only thing that does not do well in heat, almost all semiconductors start to have problems when they get hot. Typical Mosfets etc will work at up to about 120C max with some specialized ones having a max of 220C. The problem is that almost all electronic components degrade with higher heat and have a shortened life span. The big difference with String Inverters is that you can keep that heat down to a level that makes heat not an issue for the components.

As for Data on this kind of thing, I would not trust anything that comes from people in the industry. You would need a reputable independent group to do a long term study. The problem is that before you even get past a couple of years of testing the product, it has probably been replaced by a newer model. I would seriously consider Micro Inverters for a ground mount installation as Enphase seems to have excellent warranty support. Of course that warranty is not going to help you if the panels are mounted on the roof of a two story building and the one that is dead is somewhere in the middle. Your going to need a truck roll for that and it's going to be very expensive.

[solardreamer]

Of course that warranty is not going to help you if the panels are mounted on the roof of a two story building and the one that is dead is somewhere in the middle. Your going to need a truck roll for that and it's going to be very expensive.

How would an installer replace a failed string inverter at a customer site without truck roll?

BTW, lots of string inverters get installed in crowded garages without climate control or ventilation which could be even hotter than open rooftops in places that get breezes.

[robbyg]

How would an installer replace a failed string inverter at a customer site without truck roll?

BTW, lots of string inverters get installed in crowded garages without climate control or ventilation which could be even hotter than open rooftops in places that get breezes.

Comes back to the question of 24 Micro Inverters VS 1 String Inverter failing and do you care if one or two micro inverters are dead. As for replacing the sting Inverter, I would have no issue replacing my Sol-Ark 12K myself. Once installed it's pretty easy to switch off all the breakers, test with a meter and then just do a swap out. I know they are even making Inverter models now where the Top half with the inverter just lifts off and leaves the lower half with the wiring intact. As for allowing your String Inverters to overheat due to a bad installation location, that's your bad!
You have a choice in controlling the temperature around a string Inverter but you have no control with Micro Inverters.

[solar pete]

Hi All,

Give me a good string inverter everyday of the week and twice on Sundays, boxes full of electronics on roofs is a recipe for failure, we have hundreds of micro sites, hundreds of power optimiser sites and thousands of string inverter sites, I wish they were all string inverter sites

[J.P.M.]

Hi All,

Give me a good string inverter everyday of the week and twice on Sundays, boxes full of electronics on roofs is a recipe for failure, we have hundreds of micro sites, hundreds of power optimiser sites and thousands of string inverter sites, I wish they were all string inverter sites

Candor being and becoming a more scarce attribute these days, if I ever come across a guy carrying a lamp by the name of Diogenes, I'll send him in your and Solarix' direction and maybe end his quest.

[solardreamer]

Comes back to the question of 24 Micro Inverters VS 1 String Inverter failing and do you care if one or two micro inverters are dead. As for replacing the sting Inverter, I would have no issue replacing my Sol-Ark 12K myself. Once installed it's pretty easy to switch off all the breakers, test with a meter and then just do a swap out. I know they are even making Inverter models now where the Top half with the inverter just lifts off and leaves the lower half with the wiring intact. As for allowing your String Inverters to overheat due to a bad installation location, that's your bad!
You have a choice in controlling the temperature around a string Inverter but you have no control with Micro Inverters.

For DIY installers, it should be even easier to replace a microinverter or two since they installed the whole thing themselves.

[J.P.M.]

For DIY installers, it should be even easier to replace a microinverter or two since they installed the whole thing themselves.

I wouldn't confuse easy with knowledgeable. Replacing a micro in the center of an array is a PITA regardless of knowledge of how to do it. Replacing a string inverter ina garage might require moving a vehicle.

[solardreamer]

I wouldn't confuse easy with knowledgeable. Replacing a micro in the center of an array is a PITA regardless of knowledge of how to do it. Replacing a string inverter ina garage might require moving a vehicle.

Agree but I didn't say easy.

[robbyg]

For DIY installers, it should be even easier to replace a microinverter or two since they installed the whole thing themselves.

Hats off to those who did a second floor sloping roof mount by themselves. That is the one part of my system that I am getting a professional to do.
There is no way I am going up there alone to pull out several panels and change a Micro Inverter!
I have the knowledge but its safer to just spend the money.

[MGE]

Not to rain on this parade but I installed my system back in May of 2013 and have had zero problems with a single panel or micro inverter on my roof. Occasionally Once in a blue moon my Enphase Monitor Display looses network connection but I just unplug the wifi extender next to it to reboot and it all comes back up to the current time. I installed my 8.84KW DC system of (34) Suniva 260w panels with Enphase M215 micros and haven't had any issues. FWIW I like being able to view the production on their website as it lets me know when to trim certain trees if the shadows start growing and effect production.
Just checked my total lifetime production and it said 106MW. From where I sit I think it's paying out pretty well.

[bcroe]

I also still have all the original panels, which went operational in June 2013. Your lifetime
production is about 1.5 times mine per array KW, largely reflecting the constant clouds
here at 42 deg lat, and some shade. Bruce Roe

[nomadh]

This is an interesting site. I see the old hands here push hard for string inverters and it feels a little curmudgeonly while all the new sexy tech is per panel. I was sold on the per panel concept as a techie and wanted reports and all. In the end I went with sma only because at the time I had a father in law that needed a breathing machine and I have felt daytime solar was important and the extra cost of batteries was just no way. I had serious per panel envy and although have a very clear southern exposure I was sure that 3pm shade inclusion was going to destroy my output. Well I never saw any massive drop-off from shade. It seemed very linear as the sun went down
Many of the conversations pushing string inverters don't usually seem to mention shading from either side. 5 years ago shading was the primary reason to have per panel control. The word was that a few leaves or a corner of shade would DESTROY your output. Even after watching my system I was still considering adding a few tigos to affected areas.
Now I see this comparison and it shows the shading boogeyman may have been very overrated?

https://youtu.be/UQ9Szhl1ceQ

[J.P.M.]

This is an interesting site. I see the old hands here push hard for string inverters and it feels a little curmudgeonly while all the new sexy tech is per panel. I was sold on the per panel concept as a techie and wanted reports and all. In the end I went with sma only because at the time I had a father in law that needed a breathing machine and I have felt daytime solar was important and the extra cost of batteries was just no way. I had serious per panel envy and although have a very clear southern exposure I was sure that 3pm shade inclusion was going to destroy my output. Well I never saw any massive drop-off from shade. It seemed very linear as the sun went down
Many of the conversations pushing string inverters don't usually seem to mention shading from either side. 5 years ago shading was the primary reason to have per panel control. The word was that a few leaves or a corner of shade would DESTROY your output. Even after watching my system I was still considering adding a few tigos to affected areas.
Now I see this comparison and it shows the shading boogeyman may have been very overrated?
https://youtu.be/UQ9Szhl1ceQ

I don't know if it's as interesting as it is more often a no B.S. site than some others. If old hands here seem to show a preference - although I'd suggest not universal or for all applications - for string inverters, it's probably got something to do with an engineering outlook on the world gained by formal education in the sciences and engineering, and more importantly product knowledge gained by experience. Also, snowflakes be damned, some of us old farts tend to favor straight talk and see political correctness for the impediment to progress that it is.

[Ampster]

This is an interesting site. I see the old hands here push hard for string inverters and it feels a little curmudgeonly while all the new sexy tech is per panel. I was sold on the per panel concept as a techie and wanted reports and all. .......

The rules have changed the economics as well. The requirement for panel level Rapid Shut Down in many jurisdictions has made the Keep it Simple argument moot. I have had both systems and they each have their pros and cons. Visit some other forums and you will see a balanced view.